Studies on dibenz(d,f)azecine and other phenethylisoquinoline-derived alkaloids

The initial goal of this study was to develop a targetted and systematic search method for new examples of the known groups of phenethylisoquinoline-derived alkaloids with special emphasis on the least represented group, the dibenz[d,f]azecines. A considerable part of the work was devoted to the successful development of a search procedure using the combined gas chromatography-mass spectrometry (GC-MS) technique as the main tool. This included establishing working parameters for the GC-MS screening of this family of alkaloids; synthesizing some dibenz[d,f] azecines from possible new sources; and re-screening some known sources, with special attention being given to the minor components. Re-screening of Dysoxylum lenticellare, the only reported source of a naturally occurring dibenz[d,f]azecine alkaloid, along with six other species of Dysoxylum showed the absence of dibenz[d,f]azecines and other phenethylisoquinoline-derived bases. Doubts have now been raised as to the actual identity of the plant. Athrotaxis selaginoides, a homoerythrina-containing plant species, was shown to contain an extremely minor component of a dibenz[d,f]azecine base. The very low concentration did not allow its isolation. Attention was shifted to the Family Liliaceae, from which the majority of the reported phenethylisoquinoline-derived alkaloids have been isolated. Known phenethylisoquinoline-containing plants, and more than thirty other liliaceous species, were screened for dibenz[d,f]azecines. Their presence was detected only in species of Wurmbea and Baeometra. At least eight new dibenz[d,f]azecine bases were detected and one named wurmbazecine, isolated pure from Wurmbea dioica ssp. dioica. The amount of this isolate limited the n.m.r. experiments that could be done on it leaving some ambiguity in its substitution pattern. Probable structures were therefore synthesized to confirm its structure, using the eliminative ring destruction approach from tetracyclic amino ketone precursors. A biogenesis of wurmbazecine is also proposed. The conversion of the known homoerythrina alkaloid homoerythratine has led to the formation of some A-ring reduced dibenz[d,f]azecine derivatives whose mass spectral fragmentation patterns are distinct from those where the A-ring is oxidized. As a result of the alkaloid screening work, over twelve new, as well as known, wurmbaeiodeae alkaloids such as the 1- phenethylisoquinoline, homomorphines, homoaporphines, homoproaporphines, and colchicine-types from Wurmbea dioica ssp. dioica were isolated and their structures determined. Known alkaloids were also isolated from Gloriosa plantii, Wurmbea pygmaea, Baeometra uniflora and Colchicum autumnale. Other phenethylisoquinoline-type alkaloids were detected, but not isolated, from other plant species. Twelve homoerythrina-type alkaloids were also isolated and characterized from the liliaceous plants Burchardia umbellata and Burchardia mu/tiflora as well as from Athrotaxis selaginoides (Taxodiaceae). The ring D modification of the two new bases isolated from Burchardia umbellata is the first of its kind; their structures were established by spectroscopic methods. The search for the dibenz[d,f]azecine alkaloids from the family Liliaceae has allowed an in-depth study of the family's alkaloid profile. A summary of this is briefly discussed in the last discussion Chapter of the thesis, together with some possible chemotaxonomic implications